Niels Christian Nilsson is a graduate student at the Aalborg University Copenhagen, and he was at the IEEE VR presenting a poster on Factors that Effect Perceived Naturalness of Virtual Walking Speeds.
Become a Patron! Support The Voices of VR Podcast Patreon
Theme music: “Fatality” by Tigoolio
Subscribe to the Voices of VR podcast.
Rough Transcript
[00:00:05.452] Kent Bye: The Voices of VR Podcast.
[00:00:11.957] Niels Nilsson: My name is Nils, I'm a PhD student at Aalborg University Copenhagen in Denmark and I'm finishing my PhD in a couple of months and basically at the core what I've been working with is trying to see how we can make walking in place locomotion a cheap alternative to for instance the Omni or something like that. how we can make that as natural as possible. So of course the fact that you're walking in place means that the biomechanics are different from real walking and of course the proprioceptive feedback you're getting is different from real walking. So part of the early work we did was to compare different gestures for walking in place to try to see if something could actually feel more natural than walking in place. the normal gesture where you're sort of marching in the spot. And then lately we've been running studies. We demonstrated that when individuals walk on a treadmill and if you match the visual speed, for instance in the head-mounted display, to the speed of the treadmill, people tend to find those speeds too slow. And so we documented that the same is the case when you're walking in place. And now we've run a series of studies and it's starting to look like actually the degree of underestimation is a little bit different between treadmill walking and walking in place.
[00:01:18.979] Kent Bye: And why do you think that is, that there's a difference between what we're actually walking in real life versus what we are perceiving in virtual reality?
[00:01:27.473] Niels Nilsson: Okay, so if we're simply talking in virtual worlds, wearing a head-mounted display, for instance, then there could be a lot of factors that plays into the field of view being an obvious one. Compared to real-world walking, I'm not entirely sure, and actually, so from listening to some of the earlier episodes of your podcast, I think you did an interview with one of the creators of the Omni, who mentioned that he didn't have that problem, or they didn't need to apply gains, and that actually got me thinking that maybe Part of the reason is, it could be that it's sort of a bit artificial, because we perform tasks where we ask people to pay explicit attention to walking speeds, but whereas I'm assuming that when they've been calibrating the speeds they're using, they're actually having people engage in some form of gaming or whatnot, and I'm a bit curious to actually run some studies to see if that's the case, if the underestimation partially is caused by the fact that we're paying attention to the speed, We have a note about it in the paper we're presenting here, where we had a couple of subjects that came back after participating in our studies, so walking many times on a treadmill and walking in speed, or walking in place, and they came back to me after a couple of weeks saying that, you know what, it bothers me a bit, because now I've been participating in your studies, and now the speed in the real world starts to feel too slow, if I pay attention to it. So I'm a bit curious to see to what extent it's the field of view in these things, and to what extent it's the fact that we actually ask people to explicitly pay attention to it. That might be a factor as well.
[00:02:50.492] Kent Bye: Yeah, one of the things that I've been learning here is like this whole concept of expectation, and that when we're put into virtual worlds, sometimes the rules are kind of completely new. And especially if you think about the Uncanny Valley, that if it's a cartoon low-poly world versus a high-poly photorealistic world, that perhaps the more that you go towards low-poly environments that's on one spectrum of the low fidelity of the Uncanny Valley, we could have an open book in terms of what the rules are versus when you get to the high end of photorealistic then maybe things need to be more and more realistic. So I wonder if being in virtual worlds is able to kind of do this weird trip in our minds that we can kind of trick it into doing things But yeah, I don't know. And it could be from gaming or our previous expectations, the speed. But yeah, it's just this curious mismatch between the VR world and the reality. And I wonder if we'll come to a kind of a normal calibration at some point. But yeah, I don't know if you have any more thoughts.
[00:03:50.509] Niels Nilsson: No, but I think there's probably a lot of truth to that. I mean, we haven't been able... So whenever we've been running these motion perception studies where people are walking on a treadmill and walking in place, we ask them about their gaming experience, if they play first-person games. And one of the things we haven't been able to sort of find a significant difference yet, but at least from the comments, it would seem that the people that play games are inclined to like faster speeds. But again, we don't have anything to... any actual data to back it up yet. We simply haven't had enough, like, large enough samples. And I do think that expectations, as you said, play into it immensely. I think also over time, as we see VR becoming more and more widespread, I wouldn't be surprised for it to become possible to fiddle around with the speeds a lot more, and that what we perceive as natural in the real world doesn't necessarily have to apply in the virtual world at all. I suspect so, at least.
[00:04:39.167] Kent Bye: Yeah, so I know that there's a poster here that was talking about walking in place different techniques. There seemed like there's at least a handful of different techniques for walking in place. And also within the VR community, there's Matt Carroll, who is doing the stomps, which is putting IMUs on the feet to be able to detect sort of walking in place. But I'm curious, from your perspective of researching the walking in place technique, how do you actually detect that someone's marching in place? Or maybe you could sort of give an overview of the different walking in place techniques that are out there.
[00:05:09.674] Niels Nilsson: So in terms of technology, I can't give you sort of the full picture. I know there's a lot of neat techniques out there. We've used simple optical motion capture for all of ours. I think what I can comment on is the studies we've run comparing different gestures. So one of the things we found is that this traditional gesture I mentioned earlier where you're sort of marching in place and you lift your knee fairly high, people tend to find that sort of more physically straining than actually walking. And it seems like that also influences whether they find it natural or not. So we actually found that a thing that seemingly works better is to have them simply tap with their heels. So not even breaking contact with the toes, but simply stand and tap with their heels. And in that way, the amount of energy you need to spend comes closer to real walking, and people seem to like that. Of course, that also comes with some problems, because when you don't break contact with the toes, how do you turn? But I mean, I think definitely it would be worth looking more into the different types of gestures you can use for this. rather than just having people march in place. And another big problem with walking in place techniques is that they're called walking in place, but people actually don't stay in place. They tend to drift quite a lot. And we can see as well that depending on what kind of gesture you choose, the amount of drift varies quite a lot. So if walking in place techniques are actually supposed to be useful for anything, we need to figure out a way to minimize that drift.
[00:06:25.213] Kent Bye: I see. And so in your study, in your research that you were doing, what were some of the other hypotheses that you were testing and results that you're presenting here at the IEEE VR conference?
[00:06:34.313] Niels Nilsson: So what we're presenting, so last year at IEEE VR we presented two studies related to the underestimation of speeds during walking in place and we figured out that field of view influences it so of course the greater the field of view the less underestimation of speeds is occurring but there was a lot of things that we sort of, we used different displays for our past studies and we sort of saw some discrepancies between the different results and what we're presenting this year is sort of a series of follow-up studies so we've been comparing different head-mounted display weights. That doesn't seem to influence motion perception. We've been comparing the degree of peripheral occlusion, so how much can you see of the real world. We haven't found an effect of that. That might be because people, we instruct them to look straight ahead. And we've been comparing one of the things that might be a little bit, it's interesting, but also potentially problematic is that we've run a study where we compare different methods for measuring how much people underestimate or how we present the different speeds to them actually. And it turns out the methods you use also has an influence on the degree of underestimation. So it is really, I wouldn't go as far as saying that we have found sort of the exact gain you need to apply in order for a walking-in-place technique to feel natural. I think more work is needed in order to figure that out.
[00:07:42.279] Kent Bye: And is there an influence on the environment that you're walking in? Like if you're in an open field versus if there's a lot of things close to you or if there's any variation in terms of the speed that you're walking based upon the context in which you're walking?
[00:07:54.292] Niels Nilsson: So, I recall one study that explicitly addressed that. There was a workshop at VR in 2010, I think. I believe her name is Wendy Powell. They did a study where they compared two environments, but didn't find a difference. But I wouldn't rule it out as a factor. Also, one of the things we've noticed in the studies we've been running is that we always use almost an identical hallway. People are walking down, but in some cases it's had a limited length, so it's had a back wall, and in some cases it's been infinite. And it does seem like there is a bit of a difference, but it's not something we have actually carefully compared yet. And I wouldn't be surprised if there's a difference in terms of what environment you use, but I wouldn't know for sure.
[00:08:32.184] Kent Bye: And so what do you think is like the average gain in terms of what feels comfortable in walking in a virtual environment versus walking in real life then? Like is it 1.5x or 2x or what is the amount that you have to speed things up in order to feel comfortable for people?
[00:08:47.009] Niels Nilsson: So, I think after running the initial studies, the gains we've been applying when running other studies using walking in place that weren't tied to motion perception, then we've been using a gain of about 2. But again, we've been using the Enviso SX-60, which has a quite limited field of view, so I suspect if you use something like the Oculus, you'd be able to use a lower gain as well. And as we talked about earlier, if you actually have people engaged in some other activity, the gain might drop even further. And again, there might be problems with increasing the gain since it might sort of increase simulator sickness. So I think in many ways, opting to decrease the gain might be meaningful to the extent that you can.
[00:09:23.214] Kent Bye: And why do you think that is with the field of view being impacted with the judgment of how fast we're going? Do you think that there's something with our peripheral vision of being able to see things as far wide as you can that is impacting the judgment of our speed?
[00:09:37.861] Niels Nilsson: If I recall correctly, our peripheral vision is actually more sensitive to movement, which I guess makes sense from an evolutionary perspective. You would want to detect prey in the periphery, so any kind of movement might be potentially important. And also the optic flow. Basically, the optic flow is different in the periphery as well, but I suspect it is due to a higher sensitivity towards movement. But there have been many studies where people have been looking into what areas of your field of vision that are the most important. What's more important than whether it's the peripheral vision or your central field of view is the size of it. So you can also get decent sensation of motion by occluding your central vision, but having the periphery as well.
[00:10:19.977] Kent Bye: And in terms of motion sickness, what have you found in terms of navigating through space and walking in place and whether or not that tends to have less cyber sickness happening because there's some sort of actual physical movement that fools people into thinking that they're actually walking?
[00:10:35.219] Niels Nilsson: I have to pass on that one because we actually haven't been looking that much into simulator sickness, but it's definitely something that would be quite important. My impression is also in Oculus, their sort of best practice guide, that they talk a bit at least about how it's important to opt for not too fast speeds in order to minimize the risk of simulator sickness. But I don't have any sort of research we've performed that sort of backs that up. And we haven't experienced anyone talking about differences across walking in place or walking on a treadmill. But again, there's the similarity that you don't get the sensation of movement in either case from your vestibular system. So that might be why we don't see a difference.
[00:11:09.905] Kent Bye: Yeah, and one thing that I've seen and experienced myself is whenever you're changing elevation within a VR space and locomoting through it, even if you're walking in place or, you know, even just walking that, you know, some people are more sensitive to going downward or upward and I've certainly experienced that a little bit in terms of it feeling a little bit more nauseating in a VR experience when there's a change in elevation gain.
[00:11:31.592] Niels Nilsson: Yeah, I think, again, our research has been sort of quite basic, focusing on very, very simple walking tasks when comparing gestures. So following a simple flat path or walking down a corridor. So the research we've done has been basic in that sense. And the next steps will naturally be to do sort of more complicated cases such as the one you described. But I don't have a specific example there for you.
[00:11:52.378] Kent Bye: And what's next for you in terms of you're finishing up your PhD and you've been studying this walking in place, you know, what are you going to be doing next in the context of virtual reality?
[00:12:00.651] Niels Nilsson: Well, I think to the extent that I can. I'm hopefully going to stay at Aalborg University for some time still, and I'm very keen on continuing working on the whole problem of locomotion. I have yet to try the Virtuex Omni or the Virtualizer, but I'm really, really looking forward to seeing how well they work. And generally, if that's the solution, then I'll be grateful, then I'll move on to something else. Well, there might be some perceptual studies we need to do there, but anything tied to locomotion, perceptual research, user studies, and also more applied cases is what I'd like to do.
[00:12:28.690] Kent Bye: Awesome. And finally, what do you see as the ultimate potential for virtual reality and what it might be able to enable?
[00:12:35.575] Niels Nilsson: Oh, that was a big one. I think so. So it's very hard to predict because it is really sort of a completely new medium. But what I'm I can comment on what the type of VR experiences I would like to have in five years. And that is I love to be active when engaged with media. So even even though the novelty wore off, I like the Wii at first. And I think what I'm hoping is to get those immersive experiences where I can actually physically move about an environment and exploring as I would in real life.
[00:13:05.752] Kent Bye: Thanks.
